It is known that fuel level and composition information can be collected from a combination of capacitors located inside a fuel tank. The capacitance of a capacitor immersed in a fluid is proportional to the dielectric constant of the fluid times the capacitor's free air capacitance according to the following equation: EQU C.sub.r .apprxeq..epsilon..times.C.sub.a ( 1)
where:
C.sub.r =capacitance (pF) of capacitor C immersed in fluid PA1 .epsilon.=dielectric constant of the fluid in the tank PA1 C.sub.a =capacitance (pF) of capacitor C in free air PA1 C.sub.f =capacitance (pF) of capacitor C immersed in fluid PA1 .epsilon.=dielectric constant of the fluid in the tank PA1 C.sub.a =capacitance (pF) of capacitor C in free air PA1 L=total length (inches) of capacitor C PA1 H=height (inches) of the fluid in the tank
A method for sensing the level of a fluid in a fuel tank can be derived using equation (1) and the measured value of the capacitance developed on a partially immersed capacitor in a fluid of a known dielectric value. For example, U.S. Pat. No. 5,051,921, issued Sep. 24, 1991 to Paglione, describes such a method which uses interdigitated capacitors immersed in fluid contained in a fuel tank to sense the level of the fluid.
Using capacitive fuel gauges to detect the level of a fluid other than pure gasoline however, creates a problem since most capacitive fuel gauges are calibrated according to the dielectric properties of pure gasoline. An output signal from a capacitive fuel gauge immersed in a mixture of fuel ingredients, e.g. a mixture of gasoline and methanol or ethanol, registers a false reading if the gauge was calibrated to sense the level of pure gasoline since the gasoline mixture has a dielectric constant different from that of pure gasoline.
The method of the '921 patent addresses the capacitive fuel gauge calibration problem by first recognizing that the level of any fluid in a tank can be calculated according to the following equation: EQU C.sub.f .about.C.sub.a +(C.sub.a /L)[H(.epsilon.-1)] (2)
where:
As described in the '921 patent, C.sub.f, C.sub.a, L are known quantities. .epsilon. is determined using equation (1) and measured capacitance values obtained from a capacitor immersed in the fuel. The value of each of these parameters is substituted into equation (2) to determine H as follows: ##EQU1## The fuel level detection scheme disclosed in the '921 patent utilizes two capacitors connected to a monostable multivibrator, also known in the art as a "one-shot," and an oscillator or clock. The monostable multivibrator modifies the width or duration of output pulses from the oscillator or clock in proportion to the capacitance of each capacitor. The output of the first capacitor, directly proportional to the dielectric constant or composition of the fluid in the tank as indicated by equation (1), is called the composition signal. The output of the second capacitor, directly proportional to the level of the fluid in the fuel tank as indicated in equation (2), is called the level signal. An analog processor multiplies the level and composition signals together to obtain a fuel level signal which incorporates the correct dielectric constant of the fuel mixture. Thus, a signal is produced which always represents the correct fluid level regardless of fuel composition.
Although the method described in the '921 patent solves the problem of obtaining an accurate fluid level reading for mixtures other than pure gasoline from fuel gauges calibrated according to the dielectric constants of pure gasoline, it employs expensive components such as multi-vibrators, an external comparator, multiple low pass filters and an analog math co-processors which make the '921 method too expensive for use in production vehicles.
In addition, the method disclosed in the '921 patent proves disadvantageous due to the location of the composition capacitor at the bottom of the fuel tank. Water and contaminating particles inevitably find their way to the bottom of the fuel tank. The accumulation of pollutants tends to corrupt the capacitor changing the amount of charge that the capacitor stores which results in inaccurate readings.
Furthermore, the '921 patent does not provide for free-air calibration of the fuel gauge before the unit is submersed in the fluid making the '921 device very sensitive to the manufacturing tolerances of the various component parts.
Moreover, the multiple sensors used to sense level and composition data have traditionally been connected to their individual displays using individual wires. The individual wires have either been banded together into a wiring harness that exits the fuel tank through a relatively large opening or each wire has been routed out of fuel tank individually through one of many openings in the fuel tank. Using a scheme that integrates the various output data signals onto a single wire for transmission to a central vehicle computer via a multiplex bus reduces the number of openings in the fuel tank through which harmful emissions can emanate.
Finally, integration of the various functions into one fuel gauge assembly eliminates redundant subsystem components, such as multiple controllers, resulting in a significant reduction in the cost of the fuel gauge assembly and in its complexity which, in turn, facilitates the manufacturing process.